The most widely used methods to assess nerve blood flow in diabetic rats are hydrogen clearance polarography and laser Doppler flowometry, techniques requiring surgical exposure of the nerve. In these experiments, we examined the hypothesis that the trauma of surgical exposure introduces an important and hitherto largely unrecognized variable that could account for discordant reports on nerve blood flow changes induced by diabetes. We used the noninvasive (for sciatic nerve) reference sample micro-sphere method to quantify sciatic nerve blood flow in unexposed vs. surgically exposed nerves in rats with streptozotocin-induced diabetes (at different temperatures and after curarization) and in unexposed vs. surgically exposed nerves in galactose-fed rats. Baseline resting blood flow in unexposed nerves in both animal models of diabetes was either normal or increased (but was decreased in diabetic rats given d-turbocurarine). Furthermore, the normal brisk hyperemic nerve blo od flow response to the minimal trauma associated with surgical exposure of the nerve was markedly impaired in diabetic and in galactose-fed rats. Normalization of the blood flow response to trauma in galactose-fed rats by an aldose reductase inhibitor suggests that the impairment is linked to increased polyol pathway metabolism. These findings 1) confirm our previous findings that sciatic nerve blood flow in diabetic rats is increased or unchanged in unexposed nerves, while also confirming reports that in surgically exposed nerves blood flow is higher in control than in diabetic rats, and 2) indicate that blood flows in surgically exposed nerves are largely a measure of vascular responses to injury rather than (patho)physiological blood flow in undisturbed nerves. Tissue galactitol levels were measured by GC/MS and were elevated in galactose-fed animals but were normalized by aldose reductase inhibitors.